Extract domain from URL

Reduces a list of URLs to their hostnames, drops the `www.` prefix and removes duplicates — the usual first step in auditing what a page or a document links to.

Stripping `www.` is a judgement call and the right one for auditing: `www.example.com` and `example.com` are the same site to everyone except a DNS resolver, and keeping them separate makes a link audit twice as long for no benefit. Subdomains are kept, because `blog.example.com` and `shop.example.com` genuinely are different properties and often different teams. Note that the hostname is not the same as the registrable domain — reducing `blog.example.co.uk` to `example.co.uk` needs the public suffix list, which is a large data file and a different job.